“Synthetic mobility” is a technical term before it’s an exhibition title. It refers to the artificial generation of realistic human movement patterns — the simulated crowds, plausible trajectories and synthetic mobility traces that get used to train systems, model cities and stand in for populations whose real movements are either unavailable or too sensitive to use.
Synthetic Mobility, opening at Fou Gallery in New York on September 19 (reception 5–8pm) and running through November 8, takes that idea into a two-person show. It’s curated by Nina Chkareuli-Mdivani.
The pairing
Ai Makita (b. 1985, Chiba, Japan; based between Tokyo and New York) makes AI-assisted oil paintings on polyester, including Variant - b, Crowd and the 63-inch-square Variant Ⅳ, all 2026. Painting from generated source material is now common; what distinguishes the better work in this mode is what the hand does with it, and the titles here — variants, crowds — point at the show’s subject directly.
Sizhu Li (b. 1990, Sichuan, China; based in Houston) builds kinetic sculpture from aluminium, motors, lifting systems, rope and industrial fans. Works include Moonment (2024), the ongoing Pushing Hands (2024–present), and two 2026 pieces, Yuanbao and Oops, Caught’ya!.
Why the pairing is the argument
This is a better curatorial idea than “two artists who use technology.” Makita and Li are working the same problem from opposite ends.
Makita’s paintings deal with synthetic movement as image: figures and crowds that a model produced, fixed into oil paint by a hand that had to decide what to keep. Li’s sculptures deal with movement as actual physics: aluminium that a motor really lifts, fans that really push, mechanisms that wear and can fail. One asks what a generated crowd looks like; the other asks what it costs to move a real object.
Put them in one room and the gap between the two becomes the content. Simulated mobility is frictionless by definition. Motors are not.
The hyperobject frame
The curatorial text uses Timothy Morton’s hyperobjects — entities so distributed across space and time that no one can perceive them whole, climate being the standard example. It’s a heavily used concept in art writing at this point, but it’s apt here: the systems that generate synthetic movement data are exactly that kind of object. You never encounter one. You encounter its outputs, in traffic routing, in urban planning, in what a model believes a crowd looks like.
Work that renders a piece of such a system as a painting you can stand in front of, or a mechanism you can hear working, is doing the thing hyperobject-talk usually only gestures at.
Fou Gallery is at 89 Fifth Ave, Suite 701, New York.